Journal of Volcanology and Geothermal Research 369 (2019) 21–34

Contents lists available at ScienceDirect

Journal of Volcanology and Geothermal Research

journal homepage: www.elsevier.com/locate/jvolgeores

Eruption and fountaining dynamics of selected 1985–1986 high

fountaining episodes at Kīlauea volcano, Hawai'i, from quantitative
vesicle microtexture analysis
S.J. Holt a,⁎, R.J. Carey a, B.F. Houghton b, T. Orr c, J. McPhie a, S. Feig a
a
School of Natural Sciences and CODES, University of Tasmania, Private Bag 79, Hobart 7001, Tasmania, Australia
b
Department of Geology and Geophysics, University of Hawai'i, Honolulu, HI 96822, USA
c
U.S. Geological Survey, Alaska Volcano Observatory, 4230 University Drive, Suite 100, Anchorage, AK 99508, USA

a r t i c l e i n f o a b s t r a c t

Article history: Tephra from the early Hawaiian fountaining episodes of the ongoing eruption of Pu'u 'Ō'ō in the East Rift Zone
Received 11 December 2017 (ERZ) of Kīlauea provides an opportunity to study the vesicle microtextures of pyroclasts erupted from a single
Received in revised form 15 November 2018 vent over a prolonged period of time. We report the results of microtextural analysis of pyroclasts from five of
Accepted 17 November 2018
Pu'u 'Ō'ō's high (N200 m) Hawaiian fountaining episodes (episodes 32, 37, 40, 44 and 45) erupted during
Available online 19 November 2018
1985–1986. This analysis was carried out to constrain the parameters that led to large variations in fountain
Keywords:
height at Pu'u 'Ō'o, and the extent to which pyroclast residence times in the fountain modified microtextures.
Kīlauea Our results confirm the finding of Stovall et al. (2011, 2012) that pyroclasts from a single Hawaiian fountain
Hawaiian fountaining can vary greatly in texture (from bubbly to foamy), and have vesicle volume densities (Nmv) and vesicle to
Pu'u 'Ō'ō melt ratios (VG/VL) that vary by an order of magnitude. This range in vesicle texture and population is due to ex-
Vesicle microtexture tensive growth and coalescence of vesicles within the fountain after fragmentation. Only one pyroclast from four
of five episodes was found to have textures interpreted as indicative of the vesicle population near the moment of
fragmentation: bubbly texture, high density (typically N500 kg m−3), high Nmv (2.2 × 106 to 4.4 × 106), and low
VG/VL of 2.06 to 4.65. We demonstrate a linear correlation between Δ(VG/VL) and peak fountain height across a
range of Hawaiian fountains from Kilauea. This correlation could be used to infer peak heights of unobserved
Hawaiian fountaining eruptions after further testing using well-recorded events.
© 2018 Elsevier B.V. All rights reserved.

1. Introduction Hawaiian fountaining is defined as the sustained jetting of gas and

The dynamic processes within magma during its ascent in the shal-
low conduit of a volcano have a major influence on subsequent eruptive
behaviour (Gonnermann and Manga, 2013; Houghton et al., 2004;
Houghton et al., 1999; Papale, 1998; Parfitt, 2004; Vergniolle and
Jaupart, 1990). Microtextural analysis of pyroclasts is used to quantify
rates and timing of vesiculation processes in the conduit, and
fountaining dynamics during eruptions (Adams et al., 2006; Carey
et al., 2012; Costantini et al., 2010; Lautze and Houghton, 2007;
Parcheta et al., 2013; Porritt et al., 2012; Rust and Cashman, 2011;
Sable et al., 2006; Stovall et al., 2011; Stovall et al., 2012). Microtextural
datasets can be used to quantify processes during a single episode or ex-
plosion (Carey et al., 2012; Sable et al., 2006) and also to describe the
range of eruptive behaviour of a volcano during multiple eruption epi-
sodes over an extended period of time (Stovall et al., 2011; Stovall
et al., 2012).
⁎ Corresponding author.
E-mail address: sjholt@utas.edu.au (S.J. Holt).
fluidal coarse pyroclasts (Taddeucci et al., 2015; Wolff and Sumner,
2000) and is a common phenomenon at basaltic volcanoes (Fedotov
et al., 1980; Heliker et al., 2003; Polacci et al., 2006). Kīlauea Volcano
is the type-locality for Hawaiian fountaining (Fig. 1). In its written his-
torical period (since 1823), Kīlauea has had three well-documented
eruptions that produced multiple Hawaiian fountaining episodes: at
Kīlauea Iki in 1959 (Richter et al., 1970; Stovall et al., 2011; Stovall
et al., 2012), at Mauna Ulu in 1969 (Parcheta et al., 2013; Swanson
et al., 1979) and at Pu'u 'Ō'ō from 1983 to 1986 (Heliker et al., 2003;
Wolfe et al., 1987).
Hawaiian fountains constitute a low-energy end-member of
sustained (long-duration), explosive volcanism (Houghton et al.,
2016; Walker, 1973). Traditionally, the preferred model for the vesicu-
lation style and dynamics of Hawaiian-style volcanism was based on
rapid ascent of the magma within the shallow conduit, permitting me-
chanical coupling of the shallowest-exsolved and smallest bubbles of
magmatic volatile gases with the liquid melt (Cashman and Mangan,
1994; Parfitt and Wilson, 1995; Parfitt, 2004; Wilson, 1980; Wilson
and Head, 1981). However, a lack of correlation between the mass

https://doi.org/10.1016/j.jvolgeores.2018.11.011
0377-0273/© 2018 Elsevier B.V. All rights reserved.
22 S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34

Fig. 1. A) Digital Elevation Model (DEM) of Island of Hawai'i. B) DEM showing the summit and the approximate location of the upper part of the East Rift Zone of Kīlauea Volcano.

discharge rate and eruption style, as well as observed rapid fluctuations
in the eruption intensity during a single episode, show that eruptive be-
haviour is dependent not just on the rise rate of the magma, but also on
the depth, timing, and rate of volatile exsolution, and the buoyant ascent
of large vesicles within the shallow conduit (Gonnermann and Manga,
2007, 2013; Houghton and Gonnermann, 2008; Taddeucci et al.,
2015). Variables including (but not limited to) the pre-eruptive volatile
content (Johnson et al., 1994; Papale, 2005), the depth of degassing
(Edmonds and Gerlach, 2007; Spilliaert et al., 2006), bulk rheology
(Llewellin and Manga, 2005; Manga et al., 1998) and the rates of bubble
nucleation, growth, coalescence and loss (Adams et al., 2006; Cashman
and Mangan, 1994; Lautze and Houghton, 2007; Sable et al., 2006) are
complex and interrelated, and govern the explosive eruption styles of
basaltic volcanoes. It is possible to quantify aspects of the vesiculation
history, i.e., the rates of bubble nucleation, growth, coalescence and es-
cape in the magma within the shallow conduit (both pre- and post-

Fig. 2. Photographs of high Hawaiian fountains at Pu'u 'Ō'ō. 1) Incandescent, molten jet; 2) black ash and lapilli in the cooler outer region of the jet. A) Episode 45; view to the south. Peak
fountain height of episode: 257 m. Approximate fountain height at the time of photo: 200–250 m. B) Episode 44; view to the southwest. Peak fountain height of episode: 308 m. Fountain
height at the time of photo: 250–275 m.
 S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34 23

Table 1
Eruption statistics from Heliker et al. (2003) for episodes 32, 37, 40, 44 and 45 of Pu'u 'Ō'ō eruption.

Episode Sample Episode start Episode end Repose periodb Episode length Area covered by lava DRE volumec (106 Peak fountain
number (date –timea) (date –timea) (days) (days) (km2) m3) height (m)

32 04B 21/04/85 – 1516 22/04/85 – 0906 38.4 0.8 4.8 11.4 391
37 04A 24/09/85 – 1808 25/09/85 – 0619 21.8 0.5 4.4 10.3 352
40 03 01/01/86 – 1309 02/01/86 – 0238 48.5 0.6 4 8.1 264
44 02 13/04/86 – 2054 14/04/86 – 0756 22.2 0.5 5.2 8.1 308
45 01 07/05/86 – 2241 08/05/86 – 1106 23.6 0.5 5.5 6.6 257

Data from Heliker et al. (2003).

a
Hawaii-Aleutian Time Zone, UTC/GMT-10.
b
Repose period leading up to episode.
c
Dense rock equivalent volume.

fragmentation), through microtextural analysis of the erupted tephra
(Adams et al., 2006; Cashman and Mangan, 1994; Lautze and
Houghton, 2005; Mangan and Cashman, 1996; Parcheta et al., 2013;
Polacci, 2005; Sable et al., 2006; Stovall et al., 2011; Stovall et al., 2012).
The quantification of vesicularity has also been used to identify the
locations, cooling rates and residence times of pyroclasts within Hawai-
ian fountains (Porritt et al., 2012). Stovall et al. (2011, 2012) reported on
Hawaiian pyroclast morphologies coupled with qualitative and quanti-
tative microtextural evidence to show the progression of post-
fragmentation expansion during transport time within a fountain. How-
ever, these authors were able to demonstrate that clast rims with a high
vesicle number density are a valid approximation to magma conditions
pre-fragmentation. Further, Stovall et al. (2012) could demonstrate that
clasts with the highest vesicle number densities equate to the highest
intensity phase, Episode 15.
This study compares five of the later high Hawaiian fountaining ep-
isodes from Pu'u 'Ō'ō in the East Rift Zone (ERZ) in 1985–1986, which
erupted magma of nearly constant composition but had varying foun-
tain heights and durations. We quantify the vesiculation histories
through microtextural analysis of tephra erupted during episodes 32,
37, 40, 44 and 45. The microtextural characteristics of the deposits are
then linked to visual observations of fountain geometries made by the
staff of the U.S. Geological Survey's Hawaiian Volcano Observatory
(HVO). The data are used to answer three key questions: (1) do
lapilli-sized pyroclasts from these episodes retain any signature of
syn-eruption magma properties? (2) does fountain height (reflecting
decompression rate) scale with bubble number density of lapilli? and
(3) is there any quantifiable textural signature (such as vesicle-to-
melt ratio, VG/VL) that can be used as a proxy for clast residence times,
and therefore fountain height, for Pu'u 'Ō'ō fountains?

Fig. 3. A) Stratigraphic column of the deposit sampled, showing eruption episode number and sample set number for each unit. B) The location of the sample site relative to the Pu'u 'Ō'ō
vent. The photo was taken after episode 28 on December 15, 1984. Photo courtesy of HVO.
24 S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34

2. Eruption history of Pu'u 'Ō'ō vesicularity for each pyroclast was then calculated using a DRE density
of 2950 kg m−3 following Ryan (1988). Three or four pyroclasts were
The Pu'u 'Ō'ō eruption began on the 3rd of January 1983. After three chosen from each sample for microtextural image analysis. The choice
months dominated by distributed fissure fountaining, activity localised of pyroclast was a function of the density distribution, with an emphasis
at a vent later named Pu'u 'Ō'ō (Fig. 1) (Heliker et al., 2003; Wolfe to select two pyroclasts that represent the bulk of the magma erupted.
et al., 1988). From June 1983 to June 1986, 44 Hawaiian fountaining ep-
isodes took place at Pu'u 'Ō'ō. The activity was highly cyclic during this
time period, with lengthy repose periods (8 to 65 days) punctuated by
relatively short (0.3 to 16 days) Hawaiian fountaining episodes. Repose
periods were characterised by fluctuating magma level within the con-
duit and outgassing from the vent during gradual inflation of Kīlauea's
summit. Fountaining episodes consisted of low (≤200 m) to high
(N200 m) Hawaiian fountaining and lava effusion (both pāhoehoe and
'a'ā lavas), accompanied by rapid deflation at the summit of Kīlauea
(Heliker et al., 2003; Wolfe et al., 1987; Wolfe et al., 1988). This pattern
was broken in July 1986, when the eruption shifted to a new vent
(Kupaianaha), and the eruptive style changed to one of nearly continu-
ous effusion (Heliker et al., 2003).

2.1. Studied eruption episodes

The five episodes discussed in this paper (episodes 32, 37, 40, 44 and
45) took place between April 1985 and May 1986. These episodes were
chosen because all fountains exceeded 200 m in height (e.g., Fig. 2), had
products that could be precisely linked to an episode and sampled, and
were documented in detail during eruption. The visual observations are
summarised in Table 1, modified from Heliker et al. (2003). Together,
these episodes erupted a total of 4.5 × 107 m3 (Dense Rock Equivalent,
DRE) of magma and had fountain heights ranging from 264 to 391 m.
In each episode, the maximum fountain height was reached after simul-
taneous gradual build up in intensity and summit deflation. At the end
of each episode, seismic tremor within the ERZ decreased rapidly and
summit deflation switched back to inflation (Heliker et al., 2003). The
geochemical composition of the magma became steadily more mafic
through the five episodes studied and contained no traces of the early
hybrid magma that was erupted at Pu'u 'Ō'ō until Episode 30 (Garcia
et al., 1992; Moore et al., 1980).

3. Sampling and analytical methods

3.1. Field work and sample selection

A suite of samples was collected on the 5th of March 2008 from a 56-
cm-thick deposit approximately 1 km north of Pu'u 'Ō'ō that included
beds corresponding to six eruptive episodes (Fig. 3). At the time of sam-
ple collection, this location was one of very few where parts of the
1983–1986 pyroclastic succession was accessible. However, it was over-
run by lava by 2013–2016. At this location, the stratigraphy of the
tephra was described, and samples of 100 clasts in a restricted size frac-
tion (16 to 32 mm in diameter) were collected from narrow vertical in-
tervals within each fall unit following the methodology of Houghton
and Wilson (1989). These samples were used for componentry analysis,
juvenile clast density, and more detailed microtextural work. Tephra
units from most of the Pu'u 'Ō'ō episodes are not present at this site,
as the dominant direction of pyroclastic fall was southwest due to the
prevailing trade winds. Because of the infrequency of tephra fall north
of the vent, it was therefore possible to identify the episodes sampled
with a high degree of certainty (Appendix A). The deposit of Episode
30 was not studied as we were unable to obtain a sample wholly repre-
sentative of the deposit.

3.2. Analytical methods

Fig. 4. Density histograms for the pyroclast sample sets from each episode. Small numbers
The bulk density of every pyroclast within each sample was mea- within the histograms correspond to the specific pyroclasts (listed on right) chosen for
sured following the method of Houghton and Wilson (1989), measure- microtextural analysis. For comparison, blue lines are at fixed densities of 400, 800 and
ment have a ±30 kg·m−3 error, and histograms were generated. Bulk 1000 kg·m−3.
 S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34 25

Table 2
– Texture and vesicularity data for all analysed pyroclasts.

Pyroclast Episode Texturea Density (kg m−3) Density-derived vesic. (%) Nv Nmv VG/VLb

01-33 Ep. 45 B 960 67.3 5.2 × 105 1.6 × 106 2.06

01-08 F/B 840 71.4 6.1 × 105 2.2 × 106 2.50
01-06 F 620 79.0 1.5 × 105 7.3 × 105 3.76
01-01 F 460 84.6 3.9 × 104 2.5 × 105 5.49
02-04 Ep. 44 B 700 76.2 1.0 × 106 4.4 × 106 3.20
02-08 F 450 84.7 9.4 × 104 6.1 × 105 5.54
02-05 F/B 330 88.8 7.7 × 104 6.8 × 105 7.93
03-05 Ep. 40 B 770 74.1 8.9 × 105 3.5 × 106 2.86
03-07 B 520 82.3 6.8 × 105 3.8 × 106 4.65
03-17 F/B 360 87.8 7.9 × 105 6.5 × 106 7.20
03-08 F 260 91.2 3.1 × 104 3.5 × 105 10.36
04A-05 Ep. 37 B 940 68.3 1.3 × 106 4.2 × 106 2.15
04A-09 B 750 74.6 5.7 × 105 2.2 × 106 2.94
04A-01 F 330 88.7 3.7 × 104 3.2 × 105 7.85
04A-04 F 240 91.7 4.5 × 104 5.4 × 105 11.05
04B-20 Ep. 32 B 1120 62.1 3.3 × 105 8.9 × 105 1.64
04B-38 IB 870 70.7 1.7 × 105 5.7 × 105 2.41
04B-16 F 350 88.3 5.6 × 104 4.8 × 105 7.55
04B-06 F 290 90.2 5.1 × 104 5.3 × 105 9.20
a
F = foamy, B = bubbly, F/B = combination of both, IB = irregular bubbly.
b
Vesicle to melt ratio.

Additional pyroclasts were subsequently chosen near minimum and sets for each pyroclast were generated using both a Dell V105 scanner
maximum densities for samples with wide density distributions. Thin- and FEI Quanta 600 MLA Scanning Electron Microscope. Images were
sections were made from the selected pyroclasts and nested image taken at 4.5 cm, 5.0 mm and 1.25 mm horizontal field widths

Fig. 5. A) Binary image showing bubbly texture. B) Binary image showing foamy texture. C) Example of a pyroclast that exhibits both bubbly and foamy texture. Note that the transition
between the two textural domains is sharp. D) Binary image showing irregular bubbly texture.
26 S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34

Fig. 6. Four separate 500× magnification images showing the microlite species, abundance and textures from pyroclast 04B-38. In all images cpx: clinopyroxene, P: plagioclase, mag:
magnetite and ves: vesicle. See supplementary material for locations of images within the pyroclast.

(corresponding to 25×, 54× and 215× magnification) and nested to en-
sure that full ranges of vesicle sizes were effectively imaged (Shea et al.,
2010). Each nest contained from seven to twenty-one images. Adobe
Photoshop CS6 software was used to generate binary images which
were processed using the ImageJ software package (Abràmoff et al.,
2004) to measure the number of vesicles per unit area and their individ-
ual size and shape. These measurements were converted to three di-
mensions (vesicle volume density, Nv) using stereological techniques
(Sahagian and Proussevitch, 1998). To account for the presence of vesi-
cles, a melt correction is applied to the Nv values to obtain the vesicle
volume density per unit melt, Nmv (Klug et al., 2002). The vesicle-to-
melt ratio (VG/VL) was calculated for each representative pyroclast
using the following equation from Gardner et al. (1996):
 
VG Ves 1 much flatter density distribution, whereby pyroclasts with densities be-
¼
V L 1−Ves 1−m
where VG is the volume of gas (vesicles), VL is the volume of melt (glass),
Ves is the volume fraction of vesicles and m is the modal fraction of
crystals.
A detailed description of the image processing and microtextural
analysis methodology used can be found in Adams et al. (2006), Shea
et al. (2010), Stovall et al. (2011), and Parcheta et al. (2013).
4. Results
4.1. Juvenile clast density and vesicularity
Fig. 4 shows the density distributions for the pyroclast sample
sets. Three episodes, episode 32, 37 and 44 have positively skewed
density distributions; modal densities are between 300 and
500 kg m−3, corresponding to vesicularities of 89% to 82%. Of these
three episodes, the density distributions of episode 37 and episode
44 pyroclasts are unimodal, whereas the episode 32 data are bimodal
and have a secondary mode at 700 to 900 kg m−3, corresponding to
vesicularities of 74% to 67%. Episode 32 has a much larger range of
pyroclast densities (70–1330 kg m−3) than episodes 37 and 44
(240–1170 kg m−3 and 110–970 kg m−3 respectively). Although all
three sample sets are positively skewed toward low density/high ve-
sicularity, the episode 32 pyroclasts have a much larger sub-
population of pyroclasts with densities N800 kg m −3 (b71%
vesicularity).
Episode 40 pyroclasts have a modal density similar to those of epi-
sodes 32, 37 and 44 (300–400 kg m−3 and 82–89% vesicularity) but a
tween 200 and 700 kg m−3 (74% to 93% vesicularity) are of more equal
abundance. Episode 40 pyroclasts also have the narrowest density range
of the five episodes, all 100 pyroclasts having densities between 200 and
900 kg m−3 and vesicularities between 92% and 67%.
Episode 45 is the only episode that produced pyroclasts with a neg-
atively skewed density distribution. The modal density is between 800
and 900 kg m−3 (70–67% vesicularity) and less well defined than for
the three episodes that are strongly positively skewed (episodes 32,
37, 44). In addition to the highest modal density of all five episodes, ep-
isode 45 pyroclasts have the largest range, from 50 to 1350 kg m−3, cor-
responding to vesicularities between 98% and 50%. While episode 45
pyroclasts have densities mainly between 600 and 1000 kg cm−3,
there is a small secondary peak between 400 and 500 kg m−3, corre-
sponding closely to the modal densities and vesicularities of the other
four episodes.
 S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34
Fig. 7. Plot showing vesicularity versus vesicle number density (Nmv) for three high
Hawaiian fountaining eruptions at Kīlauea. This results in a b2% variation of vesicularity.
4.2. Microtextural analysis
4.2.1. Qualitative results
The pyroclasts from all five eruption episodes have vesicle
microtextures that can be classified into three distinct categories. The
texture within each studied pyroclast is indicated in Table 2. The ‘bubbly
texture’ category (‘B’ in Table 2) is dominated by small, spherical vesi-
cles (100–200 μm in diameter) that are separated by thick glass walls
(commonly 50 μm thick, but up to 100 μm) (Fig. 5A). The round shape
suggests that there has been minimal interaction between neighbouring
bubbles. Clasts with this texture are typically the least vesicular and
most dense.
The ‘foamy texture’ category (‘F’ in Table 2) is dominated by large,
closely packed polygonal vesicles, typically ≥500 μm, with bubble
walls approximately 5 μm thick (Fig. 5B). The texture of these pyroclasts
closely resembles reticulite texture as defined by Mangan and Cashman
(1996), whereby the vesicles are polyhedral and are separated by
micrometer-scale glass septa. Clasts with this texture have very high ve-
sicularities, but very few fall within the vesicularity range for reticulite,
defined as N98% by Mangan and Cashman (1996), and therefore we use
the name foamy for this textural end-member.
Although these two end-member textures differ greatly, they are not
mutually exclusive, and there is a small population of pyroclasts that ex-
hibit both textures, denoted as F/B in Table 2 (Fig. 5C). In such pyroclasts
(only 3 of the total 19 in this study), the transition between bubbly and
foamy domains is abrupt without transitional regions. Furthermore,
there is no consistent spatial distribution of the two end-member tex-
tures in a pyroclast; for example, one end-member is not confined to
the rim or core of a pyroclast.
A third vesicle microtexture, which we call ‘irregular bubbly texture’
(‘IB’ in Table 2), was found in one pyroclast from episode 32. Pyroclast
27
04B-38 has a texture similar to bubbly texture, in that the vesicles are
separated by glass walls 100–250 μm thick. However, this pyroclast
also has relatively large (1.00 and 1.24 mm) vesicles that are commonly
polylobate or irregular in shape (Fig. 5D). Pyroclast 04B-38 is the only
one within the studied sample set that contains observable microlites
(Fig. 6).
4.2.2. Quantitative results
The vesicle number densities of the measured pyroclasts (Nv; num-
ber of vesicles per unit volume of bulk rock) vary by almost two orders
of magnitude across all the episodes sampled, from 3.0 × 104 cm−3 in
episode 40 (pyroclast 03–08) to 1.3 × 106 cm−3 in episode 37 (pyroclast
04A-05) (Table 2). Clasts representing a single episode have Nv values
that span at least one order of magnitude; the largest range is in episode
37 pyroclasts, with a minimum of 3.7 × 104 cm−3 and a maximum of 1.3
× 106 cm−3. Using melt-corrected Nmv the range in vesicle number den-
sities across all the episodes is from 2.5 × 105 cm−3 in episode 45
(pyroclast 01–01) to 6.5 × 106 cm−3 in episode 40 (pyroclast 03–17)
(Table 2). The Nmv values for these five Hawaiian fountaining episodes
at Pu'u 'Ō'ō are typically less than those calculated for the 1959 eruption
of Kīlauea Iki (Stovall et al., 2011; Stovall et al., 2012) and larger than
those calculated for the 1969 eruption of Mauna Ulu (Parcheta et al.,
2013) (Fig. 7).
To better define the vesiculation histories during each eruptive epi-
sode, vesicle size distributions (VSDs) were constructed for each
analysed pyroclast (Fig. 8). These graphs show the fraction of the total
vesicle volume that is attributed to each vesicle size range, expressed
as equivalent diameter, the diameter of a circle with the same area as
calculated for the vesicle (Adams et al., 2006). In most cases the VSDs
are unimodal and have a moderate range (0.04–1.24 mm) (Fig. 8).
Within two of the five eruption episodes (episodes 45 and 37), there
is a common trend of increasing unimodality and negative skew with
respect to increasing pyroclast vesicularity for VSDs. This trend is de-
fined by the migration of the median and modal vesicle size to larger
sizes, coupled with a decrease in the number of vesicles ≤0.16 mm in
size. The VSDs of episode 32 pyroclasts have a unique trend in that the
pyroclasts have a consistent modal vesicle size between 0.4 and
0.6 mm. However, the proportion of vesicles ≤0.16 mm and ≥ 1.0 mm
decreases and the proportion of vesicles between 0.5 and 0.6 mm in-
creases, so that the modal vesicle size becomes more dominant and
the size range narrows with increasing vesicularity (Fig. 8).
5. Discussion
5.1. Determining syn-eruptive vesicle textures
Within the incandescent fountain, the cooling rates of lapilli-
sized clasts are predicted to be too slow to retain a signature of the
magma properties at the time of fragmentation (Porritt et al.,
2012). The residence time within the incandescent fountain results
in a textural evolution from a relatively dense fluidal clast to pumice,
followed by quenching during transport in a fountain (Stovall et al.,
2011, 2012). Stovall et al. (2011, 2012) showed that vesiculation
stages can be identified by comparing the Nmv of each pyroclast to
its vesicle-to-melt ratio (VG/VL) as defined by Gardner et al. (1996).
Each separate vesiculation process, such as nucleation, growth, coa-
lescence and loss (and any feasible combination of these) produces
a characteristic trend in both VG/VL versus Nmv plots and the associ-
ated VSDs (Shea et al., 2010; Stovall et al., 2011; Stovall et al., 2012).
Following Stovall et al. (2011), we interpret pyroclasts that have rel-
atively high Nmv and low VG/VL to resemble most closely the vesicu-
lation conditions of the magma at fragmentation; we call these
‘primary’ pyroclasts. For all five episodes, these primary pyroclasts
(hollow data points in Fig. 9A–C), which we interpret to approximate
the syn-fragmentation vesicle population, have bubbly textures, low
VG /VL values and high Nm v (Fig. 9), and are likely to have been
 28
S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34
Fig. 8. Histograms showing vesicle size distributions for representative pyroclasts from each high fountaining episode studied. Equivalent diameter is defined as the diameter of a circle with the same area as calculated for the vesicle (Adams et al.,
2006). Pyroclasts from each episode are arranged from top to bottom in order of increasing vesicularity (decreasing density). In each plot, the pyroclast number is given in the top right and the Nmv (bubbles cm−3), vesicularity (%) and density
(kg m−3) are listed in the top left. The vertical blue line through the 0.16 mm equivalent diameter is for comparison.
 S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34 29

Fig. 9. VG/VL versus Nmv plots for each eruption episode showing the post-fragmentation vesiculation process that is defined by the pyroclast microtextures (after Stovall et al., 2011). In all
diagrams, open circles denote primary pyroclasts and the grey arrows highlight the associated trend. The images show the microtexture of the labelled pyroclasts that are representative of
textures found for that episode. A) Episodes 37, 44 and 45, all of which show growth plus coalescence trends. B) Episode 40, which shows both a growth plus coalescence trend and a
growth plus nucleation trend. C) Episode 32, which has a growth trend. D) Key showing the possible trends and their associated post fragmentation vesicle processes. L = loss, C =
coalescence, G + C = growth plus coalescence, G = growth, G + N = growth plus nucleation, N = nucleation.

transported in the cooler outer margins of the jet (2. in Fig. 2), where
rapid cooling of the pyroclast inhibited post-fragmentation vesicula-
tion processes.
5.2. Relationship between vesicle number density (Nmv) and fountain
height
High volatile supersaturations associated with disequilibrium
(rapid) magma ascent drives the rate of bubble nucleation, as dem-
onstrated by experiments (Gardner, 2007; Le Gall and Pichavant,
2016a, 2016b) and numerical models (Burgisser and Degruyter,
2015; Gonnermann and Manga, 2013; Head and Wilson, 1989;
Parfitt, 2004; Toramaru, 1995; Vergniolle and Gaudemer, 2015;
Vergniolle and Jaupart, 1990). Thus, bubble nucleation rates, and
therefore the Nmv of the erupted tephra, can be used as a proxy for
the volatile supersaturation and decompression rate of the magma.
In the case of Hawaiian fountains, the maximum fountain height
should theoretically be directly proportional to the degree of super-
saturation and nucleation rate in the shallow ascending magma and
the N m v of the erupted tephra (Polacci et al., 2009; Stovall et al.,
2012).
In this study average Nmv for pyroclasts from each episode does not
correlate with the observed maximum fountain height (Fig. 10). In fact,
pyroclasts from the highest fountain (episode 32, 391 m) have the low-
est average Nmv (6.13 × 105 cm−3). Fig. 11 shows that the Nmv of the pri-
mary pyroclasts from Episodes 37, 40, 44 and 45 may be more closely
related to fountain height than the average Nmv for each episode, and
implies that the microtextural data from primary pyroclasts better rep-
resents the vesiculation processes that were occurring within the shal-
low conduit prior to fragmentation.
30 S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34
Fig. 10. Plot of average Nmv versus fountain height for each eruptive episode. Black bars
indicate the maximum and minimum Nmv. Numbers correspond to the episode.
Of the primary pyroclasts found, pyroclast 02–04 from episode 44
and pyroclast 01–08 from episode 45 have quenched rinds and a vesicle
size gradient from rind to interior (Fig. 11). These features suggest that,
despite both pyroclasts having the lowest VG/VL values found for their
respective episodes, even they underwent some post-fragmentation
vesicle modification. Therefore, episodes 32, 44 and 45 contain no
pyroclasts that truly exhibit pre-/syn- fragmentation microtextures, al-
though the texture of the quenched rims preserves some indications
of pre-/syn- eruptive conditions. Only pyroclasts 03–05 and 04A-05,
from the episodes 40 and 37 samples respectively, appear to be unmod-
ified by post-fragmentation vesiculation processes and preserve the
syn-fragmentation vesicle population. Both pyroclasts have bubbly tex-
tures in which larger and smaller vesicles define bands a few millime-
ters across (Fig. 11). While it is not possible to quantitatively describe
and compare the vesiculation histories and shallow conduit processes
of these high fountaining episodes based solely on these two pyroclasts,
each of their qualitative textures suggest that the ascending melt is ther-
mally and (or) mechanically heterogeneous on a small scale during
Hawaiian-style fountaining.
Due to the sparsity of primary textured pyroclasts, it is not possible,
based on Nmv, to quantitatively describe the fluid dynamics of the
magma in the shallow conduit in such a way that allows us to the
make detailed comparisons between distinct fountaining episodes at a
single vent, or more importantly, between fountaining episodes at dif-
ferent vents or different volcanoes.
5.3. Relationship between vesicle-to-melt ratio (VG/VL) and fountain height
5.3.1. Episodes 37, 44 and 45
Episodes 37, 44 and 45 (Fig. 9A) pyroclasts show trends associated
with continued growth and coalescence of vesicles after fragmentation,
most likely within the incandescent jet. For these episodes, continued
bubble growth and coalescence transforms primary bubbly textured
pyroclasts (low VG/VL, high Nmv) into mature foamy textured pyroclasts
(high VG/VL, low Nmv). Episodes 37 and 44 pyroclasts both have large
ranges in VG/VL (ranges of approximately 9 and 5, respectively),
whereas the range of VG/VL for episode 45 pyroclasts is relatively
small (a change of approximately 3.5). The comparison of these
episodes demonstrates that the magnitude of the change in VG/VL is
proportional to fountain height and thus residence time of the pyroclast
within the jet. Episodes 37 and 44 have maximum fountain heights of
352 and 308 m respectively, compared to 257 m for episode 45. We sug-
gest that lapilli remained above the glass transition temperature within
the relatively low jet of episode 45 for shorter times than during epi-
sodes 37 and 44. There was thus less time for post-fragmentation pro-
cesses to modify the vesicle population in episode 45 before
quenching, resulting in a large proportion of smaller vesicles, bubbly
textures and a negatively skewed polymodal density distribution
(Fig. 4).
Growth and coalescence of vesicles within the jet after
fragmentation results in an increase in average vesicle size, de-
crease in number density, and a thinning of vesicle walls, ulti-
mately forming closely packed polyhedral vesicles (Mangan and
Cashman, 1996). This vesicle evolution is reflected by the VSDs
for pyroclasts from each episode, most notably when comparing
episodes 45 and 37, which show the larger median and modal
vesicle sizes, coupled with fewer vesicles ≤0.16 mm in diameter,
with fountain height.
5.3.2. Episode 32
The Episode 32 pyroclasts show a simple trend in Fig. 9 with a similar
Nmv, yet large change in VG/VL. The high VG/VL pyroclasts 04B-16 and
04B-06 have foamy textures that are consistent with continued growth
of vesicles from bubbly textures (pyroclast 04B-20) in the jet after frag-
mentation (Fig. 9C). The second lowest VG/VL pyroclast, 04B-38 has a
unique texture defined by a wide range in vesicle sizes but dominated
by vesicles that are between 0.62 and 2.47 mm (Fig. 8). Large vesicles
are ragged or irregular in shape and separated by thick (100–400 μm)
glass walls (Fig. 5D).
Of the entire sample set described in this study, only pyroclast 04B-
38 from episode 32 has microlites; its glass, on average, contains 5–10%
microlites; some vesicle-poor regions contain up to 30% microlites
(Fig. 6). The qualitative textures of pyroclast 04B-38 suggest that it is,
in fact, more evolved than both the foamy textures in the high VG/VL
pyroclasts (pyroclasts 04B–16 and 04–06) and the bubbly texture of
the other low VG/VL pyroclast 04B-20. Large, irregular vesicles separated
by thick glassy regions rich in microlites are indicative of magmas that
have undergone bubble coalescence, outgassing and microlite
crystallisation prior to quenching (e.g., Polacci et al., 2006; Sable et al.,
2009; Costantini et al., 2010). The source of this sub-population of
dense, texturally evolved pyroclasts is unknown (Figs. 4, 5d). Their tex-
tures are commonly taken to reflect relatively long residence time be-
fore eruption (e.g., Cimarelli et al., 2010; Sable et al., 2006). The repose
period before episode 32 was 38.4 days and shorter than the repose pe-
riod before episode 40 (48.5 days), and yet there is no current textural
evidence that episode 40 produced observed pyroclasts with this tex-
ture. There are at least two possible sources for these texturally evolved
pyroclasts: 1) magma that stalled in one of the subterranean reservoirs
proposed by Garcia et al. (1992), or 2) lava that had previously ponded
and drained back in the conduit (Greenland et al., 1988). Whatever their
origin, it is unclear why these ragged-textured pyroclasts are observed
only in the deposits of episode 32 and not in any of the other four
episodes.
5.3.3. Episode 40
Episode 40 has the most complex vesiculation history of all five ep-
isodes. The VG/VL vs. Nmv plot suggests that, after fragmentation, some
pyroclasts underwent, bubble nucleation, growth, and coalescence
(Fig. 9B). The growth plus coalescence trend is defined by pyroclasts
03–05, 03–07 and 03–08. Pyroclasts 03–05 and 03–07 are primary
and have bubbly texture, while pyroclast 03–08 has foamy texture.
The trend defined by these two pyroclasts indicates that post-
fragmentation vesicle growth and coalescence occurred within the in-
candescent jet (similar to the processes described in Section 5.3.1).
 S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34 31

Fig. 11. A) Plot of primary pyroclast Nmv versus fountain height for each eruptive episode for which primary pyroclasts were identified. B) Qualitative microtexture of primary pyroclasts
depicted in A).
32 S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34

Pyroclast 03–17 has a high Nmv value and moderate to high VG/VL Table 3
value and appears to define a trend of growth and nucleation. However, Calculated Δ(VG/VL) values for each eruption episode.

we believe this trend to be misleading as pyroclast 03–17 has separate Pyroclast Episode VG/VL Δ(VG/VL)a
zones of foamy texture and bubbly texture, and therefore two separate 01-33 Ep. 45 2.06 3.44
vesicle size populations. 01-08 2.50
01-06 3.76
01-01 5.49
02-04 Ep. 44 3.20 4.73
5.4. Quantitative microtextural signatures as a proxy for fountain height
02-08 5.54
02-05 7.93
We define the difference between the maximum and minimum 03-05 Ep. 40 2.86 7.50
VG/VL values obtained for pyroclasts from each eruption episode as 03-07 4.65
Δ(VG/VL). Calculated Δ(VG/VL) values are used to quantify the degree 03-17 7.20
03-08 10.36
of post-fragmentation vesicle growth that took place in order to 04A-05 Ep. 37 2.15 8.89
modify pyroclasts from bubbly texture to more mature foamy tex- 04A-09 2.94
ture. This can be used to infer the residence time of a pyroclast within 04A-01 7.85
the hot plume of the fountain above the glass transition temperature, 04A-04 11.05
04B-20 Ep. 32 1.64 7.57
which in turn, is dependent on the peak fountain height (Stovall
04B-38 2.41
et al., 2011; Stovall et al., 2012). 04B-16 7.55
Table 3 and Fig. 12 show Δ(VG/VL) values for Episodes 32, 37, 40, 04B-06 9.20
44 and 45 of the 1983–1986 Pu'u 'Ō'ō eruption as well as Δ(VG/VL) a
Change in vesicle to melt ratio.
values for Episodes 1, 15 and 16 of the 1959 Kīlauea Iki eruption
from (Stovall et al., 2011; Stovall et al., 2012), plotted against the
peak fountain height for each respective episode. This plot shows Acknowledgements
that, for all eight of these selected Hawaiian fountains there is a
good linear correlation between Δ(VG/VL) and peak fountain height. We sincerely thank scientists and staff at the Hawaiian Volcano Ob-
This correlation holds true whether for discrete fountaining episodes servatory for their assistance with this research. We thank Dr. Carolyn
from a single vent or two separate eruptions from different Parcheta for her detailed review of this manuscript. We thank the anon-
vents. This correlation suggests that Δ(V G/VL ) derived from the ymous reviewers for their constructive comments and suggestions in
microtextural analysis of multiple clasts from the same narrow earlier drafts of this manuscript. This research was supported by
stratigraphic interval could be used to ascertain the approximate
peak fountain height of unobserved Hawaiian-style fountaining
eruptions.
This methodology could potentially be applied to unobserved
fountaining eruptions at Kīlauea e.g. historical caldera eruptions
(Ferguson et al., 2016; May et al., 2015) to better understand its eruptive
history and therefore its associated hazards. Furthermore, by applying
this methodology to the deposits of Hawaiian fountaining at other ba-
saltic volcanoes around the world i.e. short duration, high-intensity
lava fountaining eruptions at Mount Etna Volcano, Italy (Ganci et al.,
2012; Polacci et al., 2009) it is possible to investigate the effects of com-
plex magma properties and conduit flow regimes on the textural rela-
tionships proposed in this study.

6. Conclusions

The five high Hawaiian fountaining episodes from the Pu'u 'Ō'ō erup-
tion presented in this study had similar vesicle populations before frag-
mentation, with bubbly textures and melt- corrected vesicle volume
density (Nmv) values between 2.2 × 106 cm−3 and 6.5 × 106 cm−3.
These Nmv values fill a gap between values previously recorded for
high Hawaiian fountains at Mauna Ulu and Kīlauea Iki (Parcheta et al.,
2013; Stovall et al., 2011; Stovall et al., 2012). However, the signature
of decompression rate and magma ascent is heavily overprinted by
post-fragmentation vesiculation processes. These processes result in a
reduction in the Nmv of tephra from these episodes, except in the most
primary pyroclasts defined by low VG/VL ratio and high Nmv. Growth
plus coalescence is the dominant post-fragmentation vesiculation
process that occurs within the jet of Hawaiian fountains. Post-
fragmentation vesiculation processes have a profound effect on the
vesicle microtextures to the extent that the theoretical relationships be-
tween Nmv, explosivity and fountain height in Hawaiian-style eruptions
does not hold true. We propose that the value range between the max-
Fig. 12. Plot showing Δ(VG/VL) vs peak fountain height for all fountaining episodes
imum and minimum vesicle-to-melt ratio (Δ(VG/VL)) for the tephra of sampled in this study, as well as three episodes (1, 15 and 16) from the 1969 Kīlauea Iki
Hawaiian fountaining eruptions can be used as a proxy for peak fountain eruption from Stovall et al. (2011) and Stovall et al. (2012). Linear regression is
height. calculated on all data points.
 S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34
United States Geological Survey grant SV-ARRA-0004, Australian
Research Council grants DP110102196 and DE150101190 and National
Science Foundation, United States grants EAR-1131353 and EAR-
1521855. All SEM imaging was carried out at the Central Science
Laboratory at the University of Tasmania.
Appendix A. Sample sets were attributed to corresponding eruption
episodes using the following reference datasets
– Syn-eruption visual observations made by HVO staff members.
– HVO eruption photograph archives.
– National Oceanic and Atmospheric Administration (NOAA)
historical windspeed and direction archives.
Aerial photographic interpretation of tephra fall out by HVO staff
members preceding each eruption episode.
References
Abràmoff, M.D., Magalhães, P.J., Ram, S.J., 2004. Image processing with ImageJ. Biophoton.
Int. 11 (7), 36–42.
Adams, N.K., Houghton, B.F., Hildreth, W., 2006. Abrupt transitions during sustained ex-
plosive eruptions: examples from the 1912 eruption of Novarupta, Alaska. Bull.
Volcanol. 69 (2), 189–206.
Burgisser, A., Degruyter, W., 2015. Magma ascent and degassing at shallow levels. The En-
cyclopedia of Volcanoes, Second edition Elsevier, pp. 225–236.
Carey, R.J., Manga, M., Degruyter, W., Swanson, D., Houghton, B., Orr, T., Patrick, M., 2012.
Externally triggered renewed bubble nucleation in basaltic magma. The October 12
2008 eruption at Halema 'uma 'u Overlook vent, Kīlauea, Hawai'i, USA.
Cashman, K.V., Mangan, M.T., 1994. Physical aspects of magmatic degassing; II, Con-
straints on vesiculation processes from textural studies of eruptive products. Rev.
Mineral. Geochem. 30 (1), 447–478.
Cimarelli, C., Di Traglia, F., Taddeucci, J., 2010. Basaltic scoria textures from a zoned con-
duit as precursors to violent Strombolian activity. Geology 38 (5), 439–442.
Costantini, L., Houghton, B., Bonadonna, C., 2010. Constraints on eruption dynamics of ba-
saltic explosive activity derived from chemical and microtextural study: the example
of the Fontana Lapilli Plinian eruption, Nicaragua. J. Volcanol. Geotherm. Res. 189 (3),
207–224.
Edmonds, M., Gerlach, T.M., 2007. Vapor segregation and loss in basaltic melts. Geology
35 (8), 751–754.
Fedotov, S., Chirkov, A., Gusev, N., Kovalev, G., Slezin, Y.B., 1980. The large fissure eruption
in the region of Plosky Tolbachik volcano in Kamchatka, 1975–1976. Bull. Volcanol.
43 (1), 47–60.
Ferguson, D.J., Gonnermann, H.M., Ruprecht, P., Plank, T., Hauri, E.H., Houghton, B.F.,
Swanson, D.A., 2016. Magma decompression rates during explosive eruptions of Kī-
lauea volcano, Hawaii, recorded by melt embayments. Bull. Volcanol. 78 (10), 71.
Ganci, G., Harris, A.J., Del Negro, C., Guéhenneux, Y., Cappello, A., Labazuy, P., Calvari, S.,
Gouhier, M., 2012. A year of lava fountaining at Etna: volumes from SEVIRI. Geophys.
Res. Lett. 39 (6).
Garcia, M.O., Rhodes, J.M., Wolfe, E., Ulrich, G.E., Ho, R., 1992. Petrology of lavas from ep-
isodes 2–47 of the Puu Oo eruption of Kilauea Volcano, Hawaii: evaluation of mag-
matic processes. Bull. Volcanol. 55 (1–2), 1–16.
Gardner, J.E., 2007. Heterogeneous bubble nucleation in highly viscous silicate melts dur-
ing instantaneous decompression from high pressure. Chem. Geol. 236 (1), 1–12.
Gardner, J.E., Thomas, R.M., Jaupart, C., Tait, S., 1996. Fragmentation of magma during
Plinian volcanic eruptions. Bull. Volcanol. 58 (2–3), 144–162.
Gonnermann, H.M., Manga, M., 2007. The fluid mechanics inside a volcano. Annu. Rev.
Fluid Mech. 39, 321–356.
Gonnermann, H.M., Manga, M., 2013. Dynamics of magma ascent in the volcanic conduit.
In: Fagents, S., Gregg, T.K., Lopes, R.M. (Eds.), Modeling Volcanic Processes: the Phys-
ics and Mathematics of Volcanism. Cambridge University Press.
Greenland, L., Okamura, A., Stokes, J., 1988. Constraints on the mechanics of eruption. In:
Wolfe, E. (Ed.), The Puu O eruption of Kilauea Volcano, Episodes 1–20, January 3,
1983, to June 8, 1984. U.S. Government Printing Office.
Head, J.W., Wilson, L., 1989. Basaltic pyroclastic eruptions: influence of gas-release pat-
terns and volume fluxes on fountain structure, and the formation of cinder cones,
spatter cones, rootless flows, lava ponds and lava flows. J. Volcanol. Geotherm. Res.
37 (3), 261–271.
Heliker, C., Swanson, D.A., Takahashi, T.J., 2003. The Pu'u 'Ō'ō-Kūpaianaha Eruption of Kī-
lauea Volcano, Hawai'i: The First 20 Years. 1676. US Dept. of the Interior, US Geolog-
ical Survey.
Houghton, B., Gonnermann, H., 2008. Basaltic explosive volcanism: Constraints from de-
posits and models. Chem. Erde-Geochem. 68 (2), 117–140.
Houghton, B., Wilson, C., 1989. A vesicularity index for pyroclastic deposits. Bull. Volcanol.
51 (6), 451–462.
Houghton, B., Wilson, C., Smith, I., 1999. Shallow-seated controls on styles of explosive ba-
saltic volcanism: a case study from New Zealand. J. Volcanol. Geotherm. Res. 91 (1),
97–120.
33
Houghton, B., Wilson, C., Del Carlo, P., Coltelli, M., Sable, J., Carey, R., 2004. The influence of
conduit processes on changes in style of basaltic Plinian eruptions: Tarawera 1886
and Etna 122 BC. J. Volcanol. Geotherm. Res. 137 (1), 1–14.
Houghton, B., Taddeucci, J., Andronico, D., Gonnermann, H., Pistolesi, M., Patrick, M.R., Orr,
T., Swanson, D., Edmonds, M., Gaudin, D., 2016. Stronger or longer: discriminating be-
tween Hawaiian and Strombolian eruption styles. Geology 44 (2), 163–166.
Johnson, M.C., Anderson, A.T., Rutherford, M.J., 1994. Pre-eruptive volatile contents of
magmas. Rev. Mineral. Geochem. 30 (1), 281–330.
Klug, C., Cashman, K., Bacon, C., 2002. Structure and physical characteristics of pumice
from the climactic eruption of Mount Mazama (Crater Lake), Oregon. Bull. Volcanol.
64 (7), 486–501.
Lautze, N.C., Houghton, B.F., 2005. Physical mingling of magma and complex eruption dy-
namics in the shallow conduit at Stromboli volcano, Italy. Geology 33 (5), 425–428.
Lautze, N.C., Houghton, B.F., 2007. Linking variable explosion style and magma textures
during 2002 at Stromboli volcano, Italy. Bull. Volcanol. 69 (4), 445–460.
Le Gall, N., Pichavant, M., 2016a. Experimental simulation of bubble nucleation and
magma ascent in basaltic systems: implications for Stromboli volcano. Am. Mineral.
101 (9), 1967–1985.
Le Gall, N., Pichavant, M., 2016b. Homogeneous bubble nucleation in H2O-and H2O-CO2-
bearing basaltic melts: results of high temperature decompression experiments.
J. Volcanol. Geotherm. Res. 327, 604–621.
Llewellin, E., Manga, M., 2005. Bubble suspension rheology and implications for conduit
flow. J. Volcanol. Geotherm. Res. 143 (1), 205–217.
Manga, M., Castro, J., Cashman, K.V., Loewenberg, M., 1998. Rheology of bubble-bearing
magmas. J. Volcanol. Geotherm. Res. 87 (1), 15–28.
Mangan, M.T., Cashman, K.V., 1996. The structure of basaltic scoria and reticulite and in-
ferences for vesiculation, foam formation, and fragmentation in lava fountains.
J. Volcanol. Geotherm. Res. 73 (1), 1–18.
May, M., Carey, R.J., Swanson, D.A., Houghton, B.F., 2015. Reticulite-producing fountains
from ring fractures in Kīlauea Caldera ca. 1500 CE. In: Carey, R., Valérie, C., Poland,
M., Weis, D. (Eds.), Hawaiian Volcanoes: From Source to Surface. John Wiley &
Sons, pp. 351–367.
Moore, R.B., Helz, R.T., Dzurisin, D., Eaton, G.P., Koyanagi, R.Y., Lipman, P.W., Lockwood,
J.P., Puniwai, G.S., 1980. The 1977 eruption of Kilauea volcano, Hawaii. J. Volcanol.
Geotherm. Res. 7 (3), 189–210.
Papale, P., 1998. Volcanic conduit dynamics. In: Freundt, A., Rosi, M. (Eds.), From Magma
to Tephra: Modelling Physical Processes of Explosive Volcanic Eruptions. Elsevier Sci-
ence, pp. 55–91.
Papale, P., 2005. Determination of total H2O and CO2 budgets in evolving magmas from
melt inclusion data. J. Geophys. Res. Solid Earth 110 (B3) (1978–2012).
Parcheta, C., Houghton, B., Swanson, D., 2013. Contrasting patterns of vesiculation in low,
intermediate, and high Hawaiian fountains: a case study of the 1969 Mauna Ulu
eruption. J. Volcanol. Geotherm. Res. 255, 79–89.
Parfitt, E.A., 2004. A discussion of the mechanisms of explosive basaltic eruptions.
J. Volcanol. Geotherm. Res. 134 (1), 77–107.
Parfitt, E., Wilson, L., 1995. Explosive volcanic eruptions—IX. The transition between
Hawaiian-style lava fountaining and Strombolian explosive activity. Geophys. J. Int.
121 (1), 226–232.
Polacci, M., 2005. Constraining the dynamics of volcanic eruptions by characterization of
pumice textures. Annals of Geophysics.
Polacci, M., Corsaro, R.A., Andronico, D., 2006. Coupled textural and compositional charac-
terization of basaltic scoria: insights into the transition from Strombolian to fire foun-
tain activity at Mount Etna, Italy. Geology 34 (3), 201–204.
Polacci, M., Burton, M.R., La Spina, A., Murè, F., Favretto, S., Zanini, F., 2009. The role of syn-
eruptive vesiculation on explosive basaltic activity at Mt. Etna, Italy. J. Volcanol.
Geotherm. Res. 179 (3), 265–269.
Porritt, L.A., Russell, J.K., Quane, S.L., 2012. Pele's tears and spheres: examples from Kilauea
Iki. Earth Planet. Sci. Lett. 333–334, 171–180.
Richter, D.H., Eaton, J., Murata, K., Ault, W., Krivoy, H., 1970. Chronological Narrative of the
1959–60 Eruption of Kilauea Volcano, Hawaii. pp. 2330–7102.
Rust, A., Cashman, K., 2011. Permeability controls on expansion and size distributions of
pyroclasts. J. Geophys. Res. Solid Earth 116 (B11).
Ryan, M.P., 1988. The mechanics and three-dimensional internal structure of active mag-
matic systems: Kilauea Volcano, Hawaii. J. Geophys. Res. Solid Earth 93 (B5),
4213–4248 (1978–2012).
Sable, J.E., Houghton, B.F., Del Carlo, P., Coltelli, M., 2006. Changing conditions of magma
ascent and fragmentation during the Etna 122 BC basaltic Plinian eruption: Evidence
from clast microtextures. J. Volcanol. Geotherm. Res. 158 (3), 333–354.
Sable, J., Houghton, B., Wilson, C., Carey, R., 2009. Eruption mechanisms during the climax
of the Tarawera 1886 basaltic Plinian eruption inferred from microtextural character-
istics of the deposits. Studies in Volcanology: The Legacy of George Walker. 2. Spec.
Publ. IAVCEI, pp. 129–154.
Sahagian, D.L., Proussevitch, A.A., 1998. 3D particle size distributions from 2D observa-
tions: stereology for natural applications. J. Volcanol. Geotherm. Res. 84 (3), 173–196.
Shea, T., Houghton, B.F., Gurioli, L., Cashman, K.V., Hammer, J.E., Hobden, B.J., 2010. Tex-
tural studies of vesicles in volcanic rocks: an integrated methodology. J. Volcanol.
Geotherm. Res. 190 (3), 271–289.
Spilliaert, N., Allard, P., Métrich, N., Sobolev, A., 2006. Melt inclusion record of the condi-
tions of ascent, degassing, and extrusion of volatile-rich alkali basalt during the pow-
erful 2002 flank eruption of Mount Etna (Italy). J. Geophys. Res. 111, B04203.
Stovall, W.K., Houghton, B., Gonnermann, H., Fagents, S., Swanson, D., 2011. Eruption dy-
namics of Hawaiian-style fountains: the case study of episode 1 of the Kīlauea Iki
1959 eruption. Bull. Volcanol. 73 (5), 511–529.
Stovall, W.K., Houghton, B.F., Hammer, J.E., Fagents, S.A., Swanson, D.A., 2012. Vesiculation
of high fountaining Hawaiian eruptions: episodes 15 and 16 of 1959 Kīlauea Iki. Bull.
Volcanol. 74 (2), 441–455.
34 S.J. Holt et al. / Journal of Volcanology and Geothermal Research 369 (2019) 21–34
Swanson, D.A., Duffield, W.A., Jackson, D.B., Peterson, D.W., 1979. Chronological Narrative
of the 1969–71 Mauna Ulu Eruption of Kilauea volcano, Hawaii. US Govt. Print. Off.,
pp. 2330–7102.
Taddeucci, J., Edmonds, M., Houghton, B.F., James, M., Vergniolle, S., 2015. Hawaiian and
Strombolian eruptions. In: Sigurdsson, H. (Ed.), Encyclopedia of Volcanoes. Elsevier.
Toramaru, A., 1995. Numerical study of nucleation and growth of bubbles in viscous
magmas. J. Geophys. Res. Solid Earth 100 (B2), 1913–1931.
Vergniolle, S., Gaudemer, Y., 2015. From reservoirs and conduits to the surface: Review of
role of bubbles in driving basaltic eruptions. Hawaiian Volcanism: From Source to
Surface. 208. Am. Geophysical Union Monograph, pp. 289–322.
Vergniolle, S., Jaupart, C., 1990. Dynamics of degassing at Kilauea volcano, Hawaii.
J. Geophys. Res. 95 (B3), 2793–2809.
Walker, G.P., 1973. Explosive volcanic eruptions—a new classification scheme. Geol.
Rundsch. 62 (2), 431–446.
Wilson, L., 1980. Relationships between pressure, volatile content and ejecta velocity in
three types of volcanic explosion. J. Volcanol. Geotherm. Res. 8 (2), 297–313.
Wilson, L., Head, J., 1981. Ascent and eruption of basaltic magma on the Earth and Moon.
J. Geophys. Res. 86 (B4), 2971–3001.
Wolfe, E., Garcia, M., Jackson, D., Koyanagi, R., Neal, C., Okamura, A., 1987. The Puu O erup-
tion of Kilauea Volcano, Episodes 1–20, January 3, 1983, to June 8, 1984. In: Decker, R.,
Wright, T., Stauffer, P. (Eds.), Volcanism in Hawaii. 1. U.S. Government Printing
Office.
Wolfe, E., Neal, C., Banks, N., Duggan, T., 1988. Geologic observations and chronology of
eruptive events. US Geol. Surv. Prof. Pap. 1463, 1–97.
Wolff, J., Sumner, J., 2000. Lava fountains and their products. In: Sigurdsson, H., Houghton,
B.F., SR, McNutt, Rymer, H., Stix, J. (Eds.), Encyclopedia of Volcanoes. Academic Press,
San Diego. Academic Press, San Diego, pp. 321–329.